Constrained-size Tensorflow Models for YouTube-8M Video Understanding Challenge
arXiv:1808.06739
Abstract
This paper presents our 7th place solution to the second YouTube-8M video understanding competition which challenges participates to build a constrained-size model to classify millions of YouTube videos into thousands of classes. Our final model consists of four single models aggregated into one tensorflow graph. For each single model, we use the same network architecture as in the winning solution of the first YouTube-8M video understanding competition, namely Gated NetVLAD. We train the single models separately in tensorflow's default float32 precision, then replace weights with float16 precision and ensemble them in the evaluation and inference stages., achieving 48.5% compression rate without loss of precision. Our best model achieved 88.324% GAP on private leaderboard. The code is publicly available at https://github.com/boliu61/youtube-8m
arXiv commentAccepted paper at 2018 ECCV Youtube8M workshop: https://research.google.com/youtube8m/workshop2018/
References in corpus (14)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Distilling the Knowledge in a Neural Network
- Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
- FitNets: Hints for Thin Deep Nets
- Compressing Deep Convolutional Networks using Vector Quantization
- YouTube-8M: A Large-Scale Video Classification Benchmark
- To prune, or not to prune: exploring the efficacy of pruning for model compression
- Learnable pooling with Context Gating for video classification
- Delving Deeper into Convolutional Networks for Learning Video Representations
- Temporal Modeling Approaches for Large-scale Youtube-8M Video Understanding
- The Monkeytyping Solution to the YouTube-8M Video Understanding Challenge
- Characterizing Spatiotemporal Transcriptome of Human Brain via Low Rank Tensor Decomposition
- Deep Learning Methods for Efficient Large Scale Video Labeling
- Aggregating Frame-level Features for Large-Scale Video Classification